US 12,236,365 B2
Systems and methods for processing images to classify the processed images for digital pathology
Supriya Kapur, New York, NY (US); Christopher Kanan, Pittsford, NY (US); Thomas Fuchs, New York, NY (US); and Leo Grady, Darien, CT (US)
Assigned to Paige.AI, Inc., New York, NY (US)
Filed by PAIGE.AI, Inc., New York, NY (US)
Filed on Dec. 27, 2023, as Appl. No. 18/396,868.
Application 18/396,868 is a continuation of application No. 18/149,969, filed on Jan. 4, 2023, granted, now 11,893,510.
Application 18/149,969 is a continuation of application No. 17/705,908, filed on Mar. 28, 2022, granted, now 11,593,684, issued on Feb. 28, 2023.
Application 17/705,908 is a continuation of application No. 17/303,164, filed on May 21, 2021, granted, now 11,315,029, issued on Apr. 26, 2022.
Application 17/303,164 is a continuation of application No. 17/112,435, filed on Dec. 4, 2020, granted, now 11,042,807, issued on Jun. 22, 2021.
Application 17/112,435 is a continuation of application No. 16/875,616, filed on May 15, 2020, granted, now 10,891,550, issued on Jan. 12, 2021.
Claims priority of provisional application 62/848,703, filed on May 16, 2019.
Prior Publication US 2024/0127086 A1, Apr. 18, 2024
Int. Cl. G06N 5/04 (2023.01); G06N 20/00 (2019.01); G06T 7/00 (2017.01)
CPC G06N 5/04 (2013.01) [G06N 20/00 (2019.01); G06T 7/0012 (2013.01); G06T 2207/20076 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/30168 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method for analyzing a digital pathology image from a user, the method comprising:
receiving the digital pathology image from the user, wherein the digital pathology image is a human tissue image or an algorithmically generated image;
applying a first trained machine learning model to the digital pathology image to determine specimen property information, the first trained machine learning model having been generated by processing a plurality of digital pathology images to identify one or more parameters of the plurality of digital pathology images and predict the specimen property information associated with the plurality of digital pathology images, the training images comprising one or both of human tissue images or algorithmically generated images;
comparing predicted specimen property information to stored laboratory system information to determine if the predicted specimen property information matches the stored laboratory system information to a predetermined margin;
upon determining a mismatch to the predetermined margin between the predicted specimen property information and the stored laboratory system information, generating an alert; and
providing the alert for display via a graphical user interface of the device associated with the user.